Phenotypically distinct helper NK cells are required for gp96-mediated anti-tumor immunity

www.nature.com/scientificreports OPEN received: 21 March 2016 accepted: 23 June 2016 Published: 19 July 2016 Phenotypically distinct helper NK cells are required for gp96mediated anti-tumor immunity Abigail L. Sedlacek, Lauren B. Kinner-Bibeau & Robert J. Binder A number of Heat Shock Proteins (HSPs), in the extracellular environment, are immunogenic. Following cross-presentation of HSP-chaperoned peptides by CD91+ antigen presenting cells (APCs), T cells are primed with specificity for the derivative antigen-bearing cell. Accordingly, tumor-derived HSPs are in clinical trials for cancer immunotherapy. We investigate the role of NK cells in gp96-mediated antitumor immune responses given their propensity to lyse tumor cells. We show that gp96-mediated rejection of tumors requires a unique and necessary helper role in NK cells. This helper role occurs during the effector phase of the anti-tumor immune response and is required for T cell and APC function. Gp96 activates NK cells indirectly via APCs to a phenotype distinct from NK cells activated by other mechanisms such as IL-2. While NK cells have both lytic and cytokine producing properties, we show that gp96 selectively activates cytokine production in NK cells, which is important in the HSP antitumor immune response, and leaves their cytotoxic capacity unchanged. Select members of the heat shock protein (HSP) family are intracellular chaperones of peptides and are immunogenic1,2. Immune responses elicited by hsp703, calreticulin4, hsp905, and gp961,6 are specific for the chaperoned (peptide) antigens and have been harnessed for the immunotherapy of cancer7–9 and infectious disease10. Mechanistically, tumor-derived HSPs in the extracellular environment, as a result of extraneous administration1,3–6 or release from necrotizing cells11, engage the receptor CD91 on draining lymph node antigen presenting cells (APCs) leading to endocytosis and cross-presentation of the chaperoned peptides to T cells6,12,13. In addition, CD91 initiates signaling cascades within APCs that leads to elaboration of a panel of cytokines and up-regulation of co-stimulatory molecules11,14. As a singular entity, the HSP-peptide complex leads to priming of T cell responses and tumor rejection. The role of T cell subsets and APCs have been well defined through selective depletions of these cell types in mice15. The study of NK cells in HSP-mediated tumor rejection has been largely correlative and their role in the rejection of tumors remains vague. Immunotherapy of cancer patients with autologous, tumor derived gp96 has been shown to increase the frequency of NK cells in peripheral blood, as well as the expression of their activating receptors and IFNγfollowing re-stimulation ex vivo16,17. These effects can be enhanced by co-administration of IL-218, an NK cell activating cytokine. In a singular murine study, global deletion of NK cells was sufficient to abrogate rejection of tumors mediated by gp96 in a therapeutic setting19. The importance of these observations cannot be underestimated given the clinical application of autologous, tumor-derived HSPs for the immunotherapy of cancer7–9. Although well known for their prominent roles in viral pathogenesis, NK cells in general can be potent mediators in immunotherapy of mouse and human cancers20. As versatile lymphoid effectors, NK cells can directly recognize and lyse tumor cells21,22, but they also control the development and function of adaptive immune responses and tumor immunity23–28. In this study, the prototypical immunogenic HSP, gp96, is used to investigate the role of NK cells in rejection of tumors in murine models of cancer. Using a novel targeted mechanism of temporal depletion of NK cells, we determined that NK cells were essential during the effector phase of tumor rejection, subsequent to gp96 immunization. Although NK cells were necessary for tumor rejection following immunization with gp96, they surprisingly did not show lytic activity towards tumors. Instead, NK cells acted as helper cells and were critical for T cell-mediated tumor lysis. These NK cells were phenotypically distinct from NK cells activated via other mechanisms such as IL-2. This study provides a comprehensive assessment of the role of NK cells in HSP-mediated Department of Immunology, University of Pittsburgh, Pittsburgh, PA 15261 USA. Correspondence and requests for materials should be addressed to R.J.B. (email: ) Scientific Reports | 6:29889 | DOI: 10.1038/srep29889 1 www.nature.com/scientificreports/ Figure 1. NK cells are necessary during the effector phase of the gp96-mediated anti-tumor immune response. (A) Gp96 from tumor cells or normal mouse (non-tumor) tissue was purified to homogeneity as analyzed by SDS-PAGE. (B) Mice were immunized twice, one week apart with 2 μg of D122- or non-tumor derived gp96. Prior to tumor challenge with D122 tumor cells and throughout tumor growth mice were treated with anti-NK1.1 or control mIgG as shown. (C) Tumor growth was measured over time and is plotted as tumor area. Statistical analysis was performed by ANOVA of area under the curve followed by Bonferroni post-test *p <  0.05, **p <  0.01. tumor rejection and reveals a distinct role of these cells that exclusively displays its helper function. These studies suggest that cytotoxic and helper roles, expressed distinctly by CD8+ and CD4+ T cells respectively in the adaptive immune system, are harbored within the same cell (NK cell) in the innate immune system and are consecutively revealed by differential activation. These studies also have an impact on immunotherapy of cancer. Results NK cells are required during the effector phase of gp96-mediated tumor rejection. Homogenous tumor-derived gp96 (Fig. 1A) elicits immunity capable of rejecting a subsequent tumor challenge1,6. In this prophylactic setting, the priming phase (gp96 immunization) is distinct from the effector phase (beginning at tumor challenge) by more than a week, allowing us to specifically target the effector phase. To determine the role of NK cells in such tumor rejection, NK cells were depleted (Fig. S1) in the effector phase of the anti-tumor response, i.e., >1 week following immunization with D122-derived gp96 and 5 days before D122 tumor challenge (Fig. 1B). The delay in tumor development in D122-derived gp96 immunized mice was significantly abrogated when NK cells were depleted compared to mice treated with control IgG (Fig. 1C). Control mice immunized with normal tissue-derived gp96 exhibited no tumor rejection as previously shown. Depletion of NK cells in untreated mice did not affect normal tumor development; growth of tumors was indistinguishable from mice with a full complement of NK cells. NK cells do not play a tumor-specific cytolytic role in gp96-mediated tumor rejection. NK cells are known for their prominent role in mediating the lysis of target cells, including tumors. Given their role in the effector phase of the immune respon (...truncated)